Fake News, not just an election problem

As America moves towards another controversial election, everyone seems poised to oppose any result and point to propaganda and external parties convincing people to side one way or the other. But while elections draw a lot of attention, there remain a whole host of other services that are at risk of being overloaded, mislead or compromised by “fake news”, using basic devices and very little technical knowledge.

In our evermore digital world we keep being told that "data is king". The key commodity that is being used to make companies millions. Our personal data is bought and sold for advertisers to push more products and data is mined in every industry to show trends and predict demand. With this year's pandemic we have seen data analysis almost on a daily basis with governments basing decisions that affect millions of people on data from their science and health departments.

When looking at the impact of these decisions on entire countries, it's clear to see the importance of this data being accurate. It is this accuracy that becomes the critical factor when parsing such large datasets and it is also the biggest vulnerability when the data sources are coming from individuals.

Earlier this year there was an incident where a man in Berlin, armed with 99 mobile phones and a shopping trolley was able to cause what Google Maps thought was a huge increase in traffic, diverting anyone using the popular sat nav service to alternate routes. While this was simply an artistic exercise, it helps demonstrate what kind of an impact the invalid data plays when reviewing large datasets.

If someone can create traffic jams, this technique could also be used to divert potential customers past competitor businesses, ensuring more eyes on the perpetrator of the stunt - all without paying for billboard advertising space. The greatest strength of these crowd-sourced systems can also be the biggest weakness. With Twitter increasingly being used by news organisations to infer public opinion or activity around certain events or topics, data gets lost in the figures. None of the individual accounts can be definitively verified, with bots spouting opinion from any location on the globe and any potentially to mislead and skew the numbers.

We are now looking at a number of government tracing systems to track virus outbreaks, designed to use the crowd-sourcing strategy to ensure those exposed are notified. These apps are feeding details into the very datasets that are being used to make decisions around public health and safety, with significant consequences. There have already been a number of controversies around the efficacy of and security of these apps and datasets - not least the UK government using Excel as a database - and there is potential for the kind of sabotage that bad actors could implement to bring a country to its knees.

At a time where value placed on data has never been higher, we must ensure that the data itself is verified and accurate but most importantly, being used responsibly. Fake news isn't just on Facebook, it's in polls and statistics people work with every day and it’s the duty of the guardians of the data to keep it clean.